Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Validating Powerwall Math

This site may earn commission on affiliate links.
Hi. We have a large house in SF Bay Area with a large solar array (25 kW) that always generates excess power during the day, and quite a bit during the off-peak times when PG&E is paying us $0.03/kWH. I am trying to see if the economics pays off to add one solely on the basis of load shifting. Even on a pretty cloudy day, between 11AM and 2PM in the winter, I generate more than 15 kWH, and on sunny days in the summer, more than 80 kWH. In the evenings, we usually end up burning about 25-30 kWH during peak hours (we have EV-2A tariff).

So if we shift 24 kWH (that would be for 2 PW) from the off-peak surplus to the peak hours, that would 24 * (0.51-0.03) = $11.52 a day right? If the 2 PWs costs $19,000 (after tax break), then I would get payback after 19000/11.52/365 or 4.5 yrs. Did I calculate this right? We had wiring installed to add powerwalls later to the solar system, so I don't think the extra costs for install would be that much.

This is pretty marginal in terms of payback, though I suppose PG&E rates may end up getting even worse, and if they do maybe that will make the payback a bit better. It's a pretty marginal payback even if the cost of capital is close to zero as it is now.

Note that I am doing the PW math purely on load sifting. We have a 400A feed with a backup industrial generator (which can generate power cheaper than PG&E's peak rate), so I don't need the powerwalls for emergency power, and I don't want to redo my panels to try and have some circuits backed up and others not. The solar equipment is hooked to the utility side of the Genset ATS, so it would be easy to put the Gateway in between the main panel and the Solar panel, if I just wanted load shifting.

I assume it's possible to configure the PW to do this sort of load shifting behavior.

Am I missing something?

thx
mike
 
We had wiring installed to add powerwalls later to the solar system, so I don't think the extra costs for install would be that much.

Tesla doesnt do powerwall only installs itself, so you will be looking at third party installs, which are usually about 20k to 23k for 2 powerwalls. Perhaps you can negotiate that, but given the waiting list, I doubt it.

I also dont know if you will actually be able to install just 2 powerwalls with a PV system that large. Since you say you pre wired for it, perhaps you wired your system in such a way that you could run on half your PV + 2 powerwalls.

Otherwise, that PV is too large for 2. Since you say you are not interested in some backup and some not, if your PV cant be separated, 2 would be overloaded by that much PV.

I doubt they make sense for you.
 
  • Like
Reactions: h2ofun
We do have 3 solaredge inverters, so the solar generation can be separated out, but why don't the powerwalls just shut off when full?

Why wouldn't the system just output the excess to PG&E the way it's done now? Are you saying the PW's would have to charge using all the solar generation? I have no loads on the the panel shared with the Solar and the PW's.
 
We do have 3 solaredge inverters, so the solar generation can be separated out, but why don't the powerwalls just shut off when full?

Why wouldn't the system just output the excess to PG&E the way it's done now? Are you saying the PW's would have to charge using all the solar generation? I have no loads on the the panel shared with the Solar and the PW's.

You may not care about how it works during a power outage, but powerwalls get wired to automatically switch in a power outage. Your solar does not work now during a power outage, and because of the amount of solar and your proposed amount of powerwalls, it likely wouldnt work during a power outage when you had those powerwalls either.

You mention you dont care about it working during an outage, but I suspect that you would expect it to work the way its advertised to, and it wouldnt, with that amount of PV and that amount of solar.

When the grid was up, it would work like you expect.
 
You may not care about how it works during a power outage, but powerwalls get wired to automatically switch in a power outage. Your solar does not work now during a power outage, and because of the amount of solar and your proposed amount of powerwalls, it likely wouldnt work during a power outage when you had those powerwalls either.

You mention you dont care about it working during an outage, but I suspect that you would expect it to work the way its advertised to, and it wouldnt, with that amount of PV and that amount of solar.

When the grid was up, it would work like you expect.

Exactly - the solar would drop offline during an outage, just like it does now.

I have a diesel genset that works great for backup, doesn't depend on nat gas lines working after an earthquake, and has a 400A transfer switch that backs everything up without a fuss. The house is very efficient (all LED lights, heat pumps for AC and heating, etc...), so just a 30KW genset is more than enough to keep the house running with lots of margin, and with a 200 gal tank, it can run a long time even in bad weather without refueling.

My perfect world would have the PW's integrated with the Genset and be able to be charged by the genset and solar, and allow the generator to be run at the most efficient load, about 50% of capacity, and use solar as much as possible and the genset when solar can't handle it. But this is not possible with the way Tesla has engineered things, so my case is just on reducing the PGE burn.

So yes, the solar would not be usable during an outage, which is the way it is right now. Is this mode supported by the software in the system? There are no loads on the panel with the inverters and where the PWs would be installed.
 
The $0.03/kWh amount only applies to Net Surplus Compensation at the end of the year. In order to calculate the arbitrage savings by shifting from Off-Peak to Peak, you need to use just those $/kWh prices. Today's rates effective March 1, 2022 for PG&E EV2-A are $0.24480/kWh Off-Peak and $0.4302/kWh Peak. That is a difference of $0.1854/kWh for energy shifted during the 8 months of the Winter rate season. You should also shift the Part-Peak hours. The rate is $0.41350 Winter Part-Peak for a difference of $0.1687/kWh. Summer Peak rate is $0.55731, so the Peak differential is $0.31251/kWh from June-Sept. Part-Peak Summer differential is $0.20202/kWh. The number of kWh/day that you can shift is the lesser of your Part-Peak + Peak usage or your solar generation. I doubt you will be solar limited with that huge array unless you also have electric heat. You cannot just arbitrarily export energy to get Peak NEM credits with Powerwalls.

In order for your Powerwall system to be stable during a grid outage, you must limit the solar AC Watts to 7600 per Powerwall on the Backup side of the Gateway. One or more solar inverters can be left outside the Backup Gateway so that it will go offline when the grid goes down.

Once the Powerwalls are full and the grid is up, the Powerwalls will go into Standby, and the excess solar will go to PG&E, just like it does without Powerwalls.

A generator cannot charge Powerwalls. The generator would just be a backup when the solar+Powerwalls are exhausted. The Backup Gateway is also limited to 200A, so you would have to back up only part of your house unless you install two Gateways and 4+ Powerwalls.
 
Exactly - the solar would drop offline during an outage, just like it does now.

I have a diesel genset that works great for backup, doesn't depend on nat gas lines working after an earthquake, and has a 400A transfer switch that backs everything up without a fuss. The house is very efficient (all LED lights, heat pumps for AC and heating, etc...), so just a 30KW genset is more than enough to keep the house running with lots of margin, and with a 200 gal tank, it can run a long time even in bad weather without refueling.

My perfect world would have the PW's integrated with the Genset and be able to be charged by the genset and solar, and allow the generator to be run at the most efficient load, about 50% of capacity, and use solar as much as possible and the genset when solar can't handle it. But this is not possible with the way Tesla has engineered things, so my case is just on reducing the PGE burn.

So yes, the solar would not be usable during an outage, which is the way it is right now. Is this mode supported by the software in the system? There are no loads on the panel with the inverters and where the PWs would be installed.

I am not technical enough to say with certainty, but as you say, your generator will not charge the powerwalls (those two systems dont see each other) and I dont know what would happen in the case you specify. I know the solar would not work, I dont know how the gateway would interact with everything else since it would be expecting the powerwalls to be able to turn on the solar.

I just dont think this would work in a manner that would make sense, without having enough powerwalls to absorb the PV... but I am not a solar installer. You should consult a powerwall installer (going to have to be a third party one, and likely a really good one). The powerwalls certainly wont be "19k for 2" integrating that, I dont think, if its even possible to set it up like you want.
 
Again, I gave up on using the powerwalls for ANY backup. That's why my business case is just for load shifting, and they would be wired on the line side of the generator transfer switch, and with NO loads of any kind behind the gateway. So there is no interaction with the generator at all - effectively the solar and powerwalls are all tied to PGE, and the house loads and genset are completely isolated from them except at the main panel where the utility feed connects to the transfer switch and to the GW.

The powerwalls don't need to be stable during a power outage. If I needed to, I guess I could take one of the inverters and put it on another panel (at some expense), but can't I just tell the powerwall to drop offline during an outage? No need to charge during the outage. The gateway should see no voltage on the line side and isoilate everything behind it. It's not a real transfer switch like you would use with a generator - it just cuts off the grid right?

On the pricing, I am not assuming any export to PGE from the powerwalls (I am on NEM 2.0). The idea is to shift the excess power from solar during off-peak times to the PWs, which would then offset peak hour consumption from the house. PG&E is not paying me at the retail rates I am pretty sure based on my bill. So if I run the meter backward during off peak times, and take THAT excess power and store it in the PW, and then use it to null out any peak hour consumption from the house, the delta should be my peak hour pricing minus what PG&E would have paid me, correct? You are definitely right that peak hour pricing in the winter is 0.43 during the winter and not 0.51. But you are also right that summer peaks are now 0.55/kWH - yet another rate increase to deal with. :(

So assuming I would normally consume 24 kWH in the peak times or more per day, then I can use the stored power to offset that right? And I am just thinking about the surplus energy during off peak times, though I could also include partially peak times as well, but that would require a larger amount of storage, and I am not sure I can consume all that at night anyway.
 
I think you need to consult a powerwall installer to see if what you want is possible. The load shifting is, for sure. Your expectation that you can have the powerwalls installed in such a manner as they dont work during an outage, I have no idea. Perhaps @Vines knows.
 
If you don't use the Powerwalls for backup, the financial case for load shifting is marginal at best, even with the large rate differences on PG&E EV2-A. In a normal installation, you would get silent backup power charged with your solar that would save you on diesel running cost.

BTW, what does your annual true-up look like? Do you ever owe anything at the end of the year? I'm assuming not, based on your massive solar system. In that case, you don't get any financial savings by time shifting your solar because you have no bill to offset and TOU credits are not reimbursible.

In my case, my neighborhood had frequent power outages, I had budgeted for a nat-gas backup generator, I got a huge SGIP rebate and I owe a significant amount to PG&E at true-up every year. So, combining all those factors, the Powerwalls made clear financial sense. My bill hasn't actually gone down with the Powerwalls, but that's because PG&E changed the game by kicking me out of EV-A and giving me much lower reimbursement for my surplus solar generation. The Powerwalls allowed me to not have to pay the extra that otherwise would have happened.
 
Well, I have about 7K due in a true up next month, so there is room to improve. The gas bill is is a lot more than the lectricity though because of the pool heating (the girls like their pool warm!).

Is SGIP still available? I thought it was not an option anymore.

Maybe I should look at the using Storedge? The Lg chem batteries are pretty available, and cheaper by KWH, though they don't have inverters built in. I don't need lots of inverter power for load shifting as I am thinking about... Have any of you looked at that as an option instead of Powerwalls?
 
Every time I read one of these threads I am thankful we just have straight 1:1 production credit and flat rate kWh pricing (all in around $.22). I know it's only a matter of time before NY changes tune but I will take it while I can and hopefully be grandfathered into the credit if it should change...
 
Well, I have about 7K due in a true up next month, so there is room to improve. The gas bill is is a lot more than the lectricity though because of the pool heating (the girls like their pool warm!).

Is SGIP still available? I thought it was not an option anymore.

Maybe I should look at the using Storedge? The Lg chem batteries are pretty available, and cheaper by KWH, though they don't have inverters built in. I don't need lots of inverter power for load shifting as I am thinking about... Have any of you looked at that as an option instead of Powerwalls?
Yes, I did run the math on this one, and again, you are dealing with the fact that as a net energy exporter, kWh are priced at the wholesale rate of $0.03/kWh. There is no return on investment for any sort of reasonable timescale, and at long time scales, you are playing Russian roulette about whether any given manufacturer's batteries will have any sort of performance. Only Tesla has credible MTBF numbers, but even they don't have a credible, in my opinion, lifetime estimate. The technology is too new. Personally, I think if you don't see a reasonable rate of return in less than the warranty period, it is game over.

If you want to have them for backup, as a hedge against diesel prices, for a hedge against the inevitable earthquake, the next epidemic, or bragging rights over your neighbors, by all means, go for it. Those are use cases that have different ROIs.

But TOU arbitrage at the home level for a net energy exporter has no reasonable ROI.

BG
 
Well, I have about 7K due in a true up next month, so there is room to improve. The gas bill is is a lot more than the lectricity though because of the pool heating (the girls like their pool warm!).

Is SGIP still available? I thought it was not an option anymore.

Maybe I should look at the using Storedge? The Lg chem batteries are pretty available, and cheaper by KWH, though they don't have inverters built in. I don't need lots of inverter power for load shifting as I am thinking about... Have any of you looked at that as an option instead of Powerwalls?
you have a 7K true up with a 25 kW solar? Menlo Park does not get that hot so AC should not be a big draw. I know there are big homes in Menlo Park, but that seems like a lot. Your pool is gas heated so what is the big electrical draw?,
With a 7K true-up all of your solar back to the grid is way above $.03; at least 10x more. $.03 only applies if you are a net producer which clearly you are not.

I have 11 kW system, 4000 sq ft in Sacramento where we get 100 degrees and more often in the summer. Before PWs, my true-up was ~$200. Now with PWs, it is zero. When I did the math for rate arbitrage, I could save $900/yr (3 PWs) if I had that much clearance on my NEM, but I only had $200 available to save
 
  • Like
Reactions: h2ofun and Vines
I think you need to consult a powerwall installer to see if what you want is possible. The load shifting is, for sure. Your expectation that you can have the powerwalls installed in such a manner as they dont work during an outage, I have no idea. Perhaps @Vines knows.

It honestly sounds like the OP is trying to use the wrong product if all that is desired is load shifting.

There is a lot going on in this thread, and I will have another more complete technical response later today when some projects are done, but hardly anyone can make Powerwalls work out financially with financial load shifting alone as the value. The whole reason for the backup gateway and its switch is to provide this backup power and silently.

There is SGIP money left for large scale installations, so the OP might as well get 3 or more and get $2-3k back for each one. It's almost like buy 3 for the price of 2. SGIP has less than $20 million left though so the window is closing.

Also, very few installers (including the one I work for) will touch another installer's existing PV system and integrate Powerwalls, so that will be a challenge as well to find someone to do the work especially if you are price sensitive due to this financial payback need. The market is red hot and so with too much work to possibly do it all, installers can decline difficult jobs somewhat to focus on the best jobs where there is a good profit margin.

Construction is brutal right now with costs out of wack and material shortages. Many bottom-dollar PV installers have gone out of business. If they installed your ESS, then there is no one to service your equipment. Tesla service is a whole other conversation.
 
Note that I am doing the PW math purely on load sifting. We have a 400A feed with a backup industrial generator (which can generate power cheaper than PG&E's peak rate), so I don't need the powerwalls for emergency power, and I don't want to redo my panels to try and have some circuits backed up and others not. The solar equipment is hooked to the utility side of the Genset ATS, so it would be easy to put the Gateway in between the main panel and the Solar panel, if I just wanted load shifting.

I assume it's possible to configure the PW to do this sort of load shifting behavior.

Am I missing something?

thx
mike
Ok having reread your initial post, I see the reason you don't need backup power as you already have it. I re-iterate that you are going to overspend on backup power you say you do not need with the Tesla PW system.

However your math doesn't work out in my eyes because if you still have a bill that means you are a net consumer. If you were a net generator you might get a small check at your true up bill, and that would credit you with a tiny $0.03 per kWh.

Since you are a net consumer under NEM 2.0 your PV credits are considered at the same rate you would buy them at the time they were generated. PV power at noon generated under EV2A is worth about $0.25 per kWh as a bill offset. PV generated at 4 pm is even more valuable, though most PV systems are kicking down at this time.

The majority of the year, the best cost shift you can get is $0.18/kWh. During 4 months of the summer, you can get $0.30 delta between the lowest and highest priced power. Also Powerwalls are about 90% round trip efficient, so you need to count that into your math. My rough math says you would make less than $4 per day on the time shift of 24 kWh from off peak to peak prices.

As to the cost of the generator power, you are only looking a the fuel, not the upkeep cost per hour of runtime. Also there are some societal costs if everyone decided to fire up generators at peak power prices.

With a 400A feed and an industrial generator, you need to find a place to connect where you have a 200A subfeed, since the Gateway has this limitation.

Exactly - the solar would drop offline during an outage, just like it does now.

I have a diesel genset that works great for backup, doesn't depend on nat gas lines working after an earthquake, and has a 400A transfer switch that backs everything up without a fuss. The house is very efficient (all LED lights, heat pumps for AC and heating, etc...), so just a 30KW genset is more than enough to keep the house running with lots of margin, and with a 200 gal tank, it can run a long time even in bad weather without refueling.

My perfect world would have the PW's integrated with the Genset and be able to be charged by the genset and solar, and allow the generator to be run at the most efficient load, about 50% of capacity, and use solar as much as possible and the genset when solar can't handle it. But this is not possible with the way Tesla has engineered things, so my case is just on reducing the PGE burn.

So yes, the solar would not be usable during an outage, which is the way it is right now. Is this mode supported by the software in the system? There are no loads on the panel with the inverters and where the PWs would be installed.

If your load never exceeds 30 kW you could power your whole house on a 200A grid feed, then you could put the Tesla gateway in between the MSP and the ATS. You would have a first backup of the PV+PW and the second backup of the generator if the batteries were depleted.

To do this you either need more batteries in the backup side, or less PV in the backup side. It is possible to split the PV into 2 different systems, where 10-15 kW max of PV is on the backup side and the rest is not on the backup side. This may not be easy with your inverter and battery loations compared with the MSP.

For the size of your home and bill I'd recommend 4-5 power walls and hopefully never fire that stinky diesel generator again. Take the SGIP and the Tax credit to offset the cost and find some personal value in backup power that doesn't require combustion.

Depending on the age of the PV system, the powerwall backup system might play somewhat nice with an overabundance of PV, or it might snap on then off every 5 minutes if it's older equipment than about 3 years ago. Even newer equipment though would put your house through constant frequency shifting while on backup battery power, which isn't really ideal, but might not be a deal killer for all homes.

There is no Tesla-approved way to integrate your generator as a power source that charges the ESS, as long as your home is grid connected. Other products do offer this generator charging feature.

Again, I gave up on using the powerwalls for ANY backup. That's why my business case is just for load shifting, and they would be wired on the line side of the generator transfer switch, and with NO loads of any kind behind the gateway. So there is no interaction with the generator at all - effectively the solar and powerwalls are all tied to PGE, and the house loads and genset are completely isolated from them except at the main panel where the utility feed connects to the transfer switch and to the GW.

The powerwalls don't need to be stable during a power outage. If I needed to, I guess I could take one of the inverters and put it on another panel (at some expense), but can't I just tell the powerwall to drop offline during an outage? No need to charge during the outage. The gateway should see no voltage on the line side and isoilate everything behind it. It's not a real transfer switch like you would use with a generator - it just cuts off the grid right?

On the pricing, I am not assuming any export to PGE from the powerwalls (I am on NEM 2.0). The idea is to shift the excess power from solar during off-peak times to the PWs, which would then offset peak hour consumption from the house. PG&E is not paying me at the retail rates I am pretty sure based on my bill. So if I run the meter backward during off peak times, and take THAT excess power and store it in the PW, and then use it to null out any peak hour consumption from the house, the delta should be my peak hour pricing minus what PG&E would have paid me, correct? You are definitely right that peak hour pricing in the winter is 0.43 during the winter and not 0.51. But you are also right that summer peaks are now 0.55/kWH - yet another rate increase to deal with. :(

So assuming I would normally consume 24 kWH in the peak times or more per day, then I can use the stored power to offset that right? And I am just thinking about the surplus energy during off peak times, though I could also include partially peak times as well, but that would require a larger amount of storage, and I am not sure I can consume all that at night anyway.
This is an interesting idea, and does give what you want, but feels like a square peg in a round hole.

If you went with this approach, you might want to simply turn off the Powerwalls entirely under grid outage. There are a couple of ways to do this with the remote battery disconnect button. I have never heard of a customer not backing up any loads on any Powerwall system and we have installed a few thousand units. I suspect when you re-do the math you will see that the payback is much longer than you initially thought.
 
  • Informative
Reactions: jjrandorin
Yes, I did run the math on this one, and again, you are dealing with the fact that as a net energy exporter, kWh are priced at the wholesale rate of $0.03/kWh. There is no return on investment for any sort of reasonable timescale, and at long time scales, you are playing Russian roulette about whether any given manufacturer's batteries will have any sort of performance. Only Tesla has credible MTBF numbers, but even they don't have a credible, in my opinion, lifetime estimate. The technology is too new. Personally, I think if you don't see a reasonable rate of return in less than the warranty period, it is game over.

If you want to have them for backup, as a hedge against diesel prices, for a hedge against the inevitable earthquake, the next epidemic, or bragging rights over your neighbors, by all means, go for it. Those are use cases that have different ROIs.

But TOU arbitrage at the home level for a net energy exporter has no reasonable ROI.

BG
Wait, maybe I am confused but if he has a $7k bill with some portion of that being electricity then he is a net consumer right? Therefore NEM rules apply and he can get dollar for dollar credit for kWh generated right?
 
Thanks everyone. I have to confess I was not reading my bill properly, and the generation charges are the actual SVCE charges, not what credit I am getting. I am not close to being a net exporter. I do get a generation credit during the month, but even off peak I only exceed the what I use a few months in the summer, and I never get close to offsetting peak hour use. The bill is not that easy to understand, but I will dig into it more.

So you are clearly right about the offset of load shifting being a lot smaller than what I was assuming, so the PW's clearly don't pencil out on the load shifting side.

Part of the reason electricity usage is high is I have a couple racks of servers in the basement that I use for a home lab. On average it's about 1800 watts, but it is continuous load - disks spinning, network switches, etc... Still, that's not explaining why in the middle of the night with the lights out I still am burning 5-6 kilowatts every hour. Pool pumps and car chargers shouldn't be running then, so something more investigation is called for.

In any case, the powerwalls clearly don't make sense, but I appreciate all the help I got here! Thanks everyone!
 
  • Like
Reactions: h2ofun